Work & Productivity

About 35% of American Workers Did Some of Their Work at Home in 2025

By Marcus Delgado / November 6, 2026

Woman working on a laptop at a table in a sunny room at home

Quick answer: About 35% of employed Americans did some or all of their work at home on the days they worked in 2025, according to the Bureau of Labor Statistics’ American Time Use Survey, released June 25, 2026. That is not the same as “35% of workers are remote.” It counts anyone who did even part of a workday’s work at home, and about 70% did some of it at their workplace, so lots of people are in both groups. As of November 8, 2026, it is the latest American Time Use Survey release.

Key takeaways

The exact measure: the share of employed people, among those who worked on the survey day, who did some or all of that work at home. BLS’s Table 6 shows 34.5%, which the release text rounds to 35%.
Women (38.2%) were more likely than men (31.3%) to work at home, and bachelor’s degree holders (51.4%) far more than people with only a high school diploma (19.0%).
2025 sits near the 2023 level (35%) and well above 2019 (24%), but a government shutdown left a hole in the 2025 data.
Other surveys measure other things. WFH Research reports the share of full paid days worked from home, about 27% in October 2026: a different yardstick, not a conflict.

What the 35% actually measures

The American Time Use Survey (ATUS) is not a headcount of remote jobs. Interviewers ask a sample of Americans to walk through the day before the interview, activity by activity, including where they were. BLS then averages those diaries across the year. In 2025, about 6,100 people were interviewed.

From that, BLS builds Table 6, which covers employed people who worked on their diary day. In 2025 that was 65.5% of the 167.4 million employed people age 15 and over, or roughly 109.7 million on an average day. Of those workers, 34.5% (about 37.8 million) did some work at home and 69.8% did some work at their workplace.

Those add up to more than 100% because, as the table’s footnote says, a person may work in more than one place in a day. The footnotes add two more details. “Working at home” means any time spent working at home, and it is not limited to people whose usual workplace is their home. The work counted covers main and other jobs and excludes work-related travel.

BLS counts someone as employed if they did any paid work, or any work in their own business, in the seven days before the interview.

The 2025 numbers at a glance

These figures come from BLS’s 2025 annual averages (Table 6), among employed people who worked on an average day.

Group	Worked at home	Worked at workplace
All employed, 15 and over	34.5%	69.8%
Full-time workers	34.9%	70.3%
Men	31.3%	72.2%
Women	38.2%	67.1%
Age 25+, high school diploma, no college	19.0%	83.1%
Age 25+, some college or associate degree	29.7%	73.7%
Age 25+, bachelor’s degree or higher	51.4%	56.2%

Education is the standout gap. Among workers 25 and older, the share working at home runs from 9.1% for people without a high school diploma to 56.7% for people with advanced degrees. On hours, those who worked at home averaged about 5.1 hours of it that day, versus about 7.9 hours for those at their workplace.

Compared with earlier years and other surveys

BLS’s earlier releases give the same measure for other years, using the rounded figures in each release text:

2019: 24% (BLS later published the unrounded figure as 23.7%).
2023: 35% (34.6% unrounded).
2024: 33% in the release text, 32.5% in its Table 6.
2025: 35% in the release text, 34.5% in Table 6.

So the short version is a jump after 2019 and a plateau in the low-to-mid 30s since. The two-point step from 2024 to 2025 is small, and I did not find BLS saying whether it is statistically meaningful. Treat it as “roughly flat,” not “remote work is rising again.”

Other sources use different definitions, so their numbers are not interchangeable:

Source	What it counts	Recent number
BLS American Time Use Survey	Employed people who did some or all of their work at home on a day they worked	34.5% in 2025
WFH Research, “Work from Home Rate”	Percent of full paid working days done from home	27.06% in October 2026
BLS Current Population Survey	People who teleworked or worked at home for pay at any time in the survey week, as a share of those at work	Not used here

WFH Research’s data file (released at the end of October 2026) shows monthly readings of about 24.5% to 27.3% from November 2025 through October 2026. Counting days instead of people is why it looks lower.

What it does not mean

It does not mean one in three people is fully remote

Someone who answered email from the couch for an hour after a day at the office counts in the “worked at home” group. BLS’s own explainer for earlier data says the average includes all time spent working at home, whether or not it was on a scheduled workday.

It does not tell you about hybrid schedules

The survey sees one day per person, so it cannot say how many days a week anyone is at home.

It does not say whether remote work is good or bad

ATUS measures where time was spent. It says nothing about productivity, pay or what workers want.

The shutdown asterisk

The federal government shutdown ran from October 1 to November 12, 2025. ATUS suspended operations during that window, with no calls and no completed interviews, so there are no diary data covering September 30 through November 11. BLS weighted the diaries it did collect from November 12 through December 30 to stand in for the whole fourth quarter, and it says it is not possible to quantify the effect on the estimates. Any 2025 versus 2024 comparison carries that caveat.

A quick checklist before you quote this number

Say “on days they worked” and name the population: employed people 15 and over.
Include the year (2025) and the source (BLS American Time Use Survey).
Use “did some or all of their work at home,” not “worked remotely.”
Mention that about 70% also did some work at their workplace.
Decide whether to use 35% (release text) or 34.5% (Table 6) and stay consistent.
Do not compare it to a days-based measure such as WFH Research without saying they differ.
Note the 2025 shutdown gap if you compare with other years.

Frequently asked questions

Is 35% the share of jobs that are remote?

No. It is the share of employed people, among those who worked on the survey day, who did some or all of their work at home. It does not count jobs, and it does not classify anyone as remote or hybrid.

Why do other surveys give a lower number, like 27%?

They count different things. WFH Research reports the share of full paid days done from home, about 27% in October 2026. A share of days will usually be lower than a share of people who did any work at home.

Did the 2025 figure go up from 2024?

The release text says 33% for 2024 and 35% for 2025, and Table 6 shows 32.5% and 34.5%. BLS has said it cannot quantify the shutdown’s effect on 2025 estimates, so I would call it roughly stable.

Will there be a newer number soon?

Not from ATUS, which has published its annual results in late June in recent years. WFH Research updates its data monthly.

Before you decide

If you are writing a report, post or slide, the safest line is: in 2025, about 35% of employed people did some or all of their work at home on the days they worked, per BLS. Add a clause on what that counts and you have prevented the most common misreading.

If you are deciding something personal, like taking a job or setting a team policy, a national average is a weak guide. Occupation matters more, and the education gap above shows why. Ask the employer what the actual in-office expectation is.